Journal of Abnormal Psychology
2002, Vol. 111, No. 2, 302–312
Copyright 2002 by the American Psychological Association, Inc.
0021-843X/02/$5.00 DOI: 10.1037//0021-843X.111.2.302
Crying Threshold and Intensity in Major Depressive Disorder
Jonathan Rottenberg, James J. Gross, Frank H. Wilhelm, Sadia Najmi, and Ian H. Gotlib
Stanford University
Clinical lore suggests that depression is associated with frequent and intense crying. To test these
postulations empirically, a standardized cry-evoking stimulus was presented to depressed and nondepressed participants, and their likelihood of crying and the magnitude of crying-related changes in their
emotion experience, behavior, and autonomic physiology were compared. Unexpectedly, crying was no
more likely in depressed than in nondepressed participants. Within the nondepressed group, participants
who cried exhibited increases in the report and display of sadness and had greater cardiac and
electrodermal activation than did participants who did not cry. There was less evidence of this
crying-related emotional activation within the depressed group. The lack of emotional activation among
clinically depressed participants who cried provides a tantalizing clue concerning how emotions are
dysregulated in this disorder.
examine the impact of depression on the likelihood of crying and
the intensity of the crying response.
Major depressive disorder (MDD) is commonly referred to as a
disorder of emotion (e.g., Barlow, 1988; Gross & Muñoz, 1995).
Indeed, the centrality of emotional disturbance is readily apparent
from the diagnostic features of MDD. Major depression, as described in the fourth edition of the Diagnostic and Statistical
Manual of Mental Disorders (DSM–IV; American Psychiatric Association, 1994), is characterized by persistent sad mood or a loss
of interest or pleasure in daily activities and by a number of
associated symptoms, such as weight loss or gain, loss of appetite,
sleep disturbance, psychomotor retardation, fatigue, and feelings
of guilt. Given the range of emotional disturbance in depression, it
is surprising that scientists have only recently begun to study and
specify precisely how emotion is affected in MDD (see Kring &
Bachorowski, 1999, for a review).
One window into the emotional functioning of depressed individuals is provided by the crying response. Crying has long been
viewed as a universal response involved in the communication of
distress (Darwin, 1872), and clinicians have often remarked that
depressed persons cry excessively (cf. Beck, Rush, Shaw, & Emery, 1979). The DSM–IV, however, does not require changes in
either the threshold for crying or the intensity of the crying
response for a diagnosis of MDD, and the significance of these
oft-repeated clinical observations is unclear. In the following sections, we review the empirical literature on crying in nondepressed
and depressed individuals and then describe a study designed to
Crying in Nondepressed Individuals
Crying is a secretomotor response defined by the shedding of
tears from lacrimal structures. Crying while sad is often accompanied by grimacing and other facial expressions of sadness, by
vocalizations, and by the convulsive inhalation and exhalation of
air. Like the solicitation signals of other species (Maclean, 1985;
Newman, 1985), crying in human infants has been conceptualized
as a signal that evolved to increase the proximity between infants
and their caregivers (Ainsworth, 1969; Bowlby, 1969; Hunziker &
Barr, 1986). Although crying in adults is undoubtedly more complex than crying in infants, like infant crying, adult crying has been
interpreted as a signal of distress (Cornelius, 1984, 1988). Indeed,
theorists have argued that crying is designed to alleviate distress by
motivating the self to action (Tompkins, 1963), by motivating
others to engage in prosocial behaviors (Averill, 1968), or by
furthering physiological homeostasis (Efran & Spangler, 1979;
Frey, DeSota-Johnson, Hoffman, & McCall, 1981).
In keeping with the view that crying is an aversive, behaviorally
activated response that signals distress, empirical investigations
have demonstrated that crying leads to coordinated changes across
systems of emotional responding. For example, individuals who
display tears while viewing a sad film report more sadness and
pain and exhibit more expressive behavior than do their noncrying
counterparts (Gross, Fredrickson, & Levenson, 1994). Adults who
cry during a sad film also have significantly greater increases in
sympathetic nervous system activity than do adults who simply
watch a sad film, as evidenced by increases in measures of electrodermal response (Kraemer & Hastrup, 1988), decreases in peripheral temperature, and accelerations in heart rate beyond the
demands of ongoing somatic activity (Gross et al., 1994).
Jonathan Rottenberg, James J. Gross, Sadia Najmi, and Ian H. Gotlib,
Department of Psychology, Stanford University; Frank H. Wilhelm, Department of Psychiatry and Behavioral Sciences, Stanford University.
This research was supported by National Institute of Mental Health
(NIMH) Grant MH-59259 awarded to Ian H. Gotlib, NIMH Grant MH581477 awarded to James J. Gross, and an NIMH predoctoral fellowship in
affective science awarded to Jonathan Rottenberg.
We express our appreciation to Jessica Bernstein, Megan McCarthy, and
Kate Reilly for their help in conducting this study.
Correspondence concerning this article should be addressed to Ian H.
Gotlib, Department of Psychology, Stanford University, Jordan Hall,
Building 420, Stanford, California 94305-2130. E-mail: gotlib@psych
.stanford.edu
Crying in Depressed Individuals
Textbook descriptions of depressed patients often include increased frequency of crying as a characteristic of the disorder (e.g.,
302
CRYING IN DEPRESSION
Beck, 1967). The self-reports of depressed individuals on questionnaire or interview items (e.g., Beck et al., 1979; Hamilton,
1960) also suggest that crying may be increased in this disorder.
This makes good sense: A common conception of crying behavior
in adults is that it registers distress. Depressed individuals are well
known to report acute negative affect (Watson, Clark, & Carey,
1988) and high levels of personal distress (American Psychiatric
Association, 1994). More specifically, the emotional state that
most immediately precedes crying among nondepressed individuals is feeling sad (Wallbott & Scherer, 1988). Because sad mood is
a defining feature of MDD, it seems reasonable to expect that the
depressed state lowers the threshold for crying behavior and magnifies the coordinated changes in emotion that typically occur
during crying. Interestingly, recent neurobiological accounts of
mood disorders also predict these crying-related changes in depression, describing pathophysiology whereby the same level of
eliciting stimulus evokes increasing behavioral and physiological
reactivity over time (see Post & Weiss, 1998, for a review of
kindling and behavioral sensitization models).
Existing research, however, does not yet provide strong evidence that depression fundamentally alters either the frequency of
crying or the magnitude of the crying response. Indeed, previous
investigations examining the association between crying and depressed mood in nonclinical samples have yielded largely inconclusive results. For example, whereas Frey et al. (1981) found a
positive association between frequency of crying and self-ratings
of depression, but only for females, Choti, Marston, Holston, and
Hart (1987) reported the opposite pattern of results: Baseline
reports of depression were related to reports of crying in response
to a film stimulus, but only for males. Similarly, Hammen and
Padesky (1977) reported that elevated scores on the Beck Depression Inventory were associated with a reported inability to cry, but
only for males. In other studies, frequency of crying was found to
be unrelated to self-report measures of depressed mood (e.g.,
Kraemer & Hastrup, 1986; Labott & Martin, 1987).
It is possible that a more robust association between depression
and crying would be found if diagnosably depressed individuals
were studied or if behavioral measures of crying were used.
Unfortunately, prior studies of crying in clinically depressed samples have been limited to describing demographic and clinical
features associated with this behavior. For example, on the basis of
ward observations, Davis, Lambert, and Ajans (1969) concluded
that crying was more common among neurotic than among psychotic depressives. Although some investigators have found depressed females to be more likely than depressed males to report
increased crying (e.g., Carter, Joyce, Mulder, Luty, & McKenzie,
2000; Vredenburg, Krames, & Flett, 1986; Zetin, Sklansky, &
Cramer, 1984), there is evidence to suggest that this gender difference may be less reliable in depressed samples than it is among
nondepressed controls (e.g., Choti et al., 1987; Davis et al., 1969;
Kraemer & Hastrup, 1986; Williams, 1982). Importantly, none of
these prior investigations have applied a theoretical framework
regarding the known functions of crying to depression or have
performed a controlled assessment of crying threshold. It is clear,
therefore, that more research is required to understand the impact
of depression on both the frequency of crying and the intensity of
the crying response.
303
The Present Study
The present study was designed to examine the crying response
in depression. More specifically, we tested three interrelated
hypotheses.
Hypothesis 1: Compared with nondepressed controls, depressed
participants will be more likely to cry in response to a sad
stimulus. To evaluate this hypothesis, we videotaped clinically
depressed and nondepressed participants while they viewed an
emotionally provocative, cry-eliciting film clip. Having participants view this sad film clip provided an ethical, controlled, and
replicable means for us to assess crying threshold in depression.
Previous work with comparable stimuli indicated that only
about 20% of an unselected sample would be expected to meet our
strict behavioral criterion for crying (Gross et al., 1994); consequently, we enrolled a large sample of participants (N ⫽ 104) into
this study. This behavioral criterion of observable tears also provided us with an objective measure of crying, thereby allowing us
to circumvent the biases in reporting and memory that characterize
depressed individuals (e.g., Gotlib & Neubauer, 2000). We predicted that depressed participants would be more likely to exhibit
tears during the sad film than would nondepressed participants. To
test this hypothesis, we used the chi-square statistic to compare the
proportion of depressed and nondepressed participants who cried
in response to the film.
Hypothesis 2: Depression will enhance the crying response.
Prior work with nondepressed individuals indicates that crying
leads to increases in the report and expression of sadness and in
physiological arousal (Gross et al., 1994). We predicted that depression would further enhance this crying response. To evaluate
this hypothesis, we first divided each diagnostic group into criers
and noncriers (on the basis of the presence or absence of observable tears). Then, within each diagnostic group, we compared
criers and noncriers on measures that indexed the intensity of
emotional responding under neutral and sad conditions. We predicted that differences between criers and noncriers in their reports
and displays of sadness and in their sympathetic physiology would
be greater within the sample of depressed participants than would
be the case within the nondepressed sample.
In conducting our analyses to test this hypothesis, the primary
between-subjects variables were depression status (depressed, nondepressed) and cry status (crier, noncrier), and the within-subject
variable was film condition (neutral, sad). Gender was also included as a between-subjects factor. Finally, medication use (medicated, unmedicated) was included in the analyses as a covariate.1
1
Because 31 of the 71 depressed participants were receiving pharmacotherapy, we conducted a number of analyses to examine possible medication effects. First, we compared depressed participants who were taking
psychotropic medication with those who were not taking medication on
their baseline physiology. A multivariate analysis of variance revealed that
medication use was not related to baseline levels of sympathetic nervous
system variables, F(1, 66) ⬍ 1. Because hypotheses concerned reactivity to
a sad film, it was important to examine the impact of medication on levels
of experiential, behavioral, and physiological reactivity. There were no
significant differences between depressed participants on medication and
depressed participants not on medication with respect to changes in selfreported and observer-rated sadness, both ts(70) ⬍ 1, nor was there a
significant effect of medication on physiological responses to the sad film,
ROTTENBERG, GROSS, WILHELM, NAJMI, AND GOTLIB
304
If the effects of crying in response to a sad stimulus were enhanced
in depressed individuals, we would obtain significant interactions
of cry status, film condition, and diagnostic group for each emotion response domain. We could then test our hypothesis by
decomposing this interaction and examining the relative magnitude of differences as a function of film condition between criers
and noncriers within each diagnostic group. We conducted separate analyses on self-reported sadness, positive affect, and negative
affect, observer-rated sadness behavior, and six measures of autonomic functioning.
Hypothesis 3: Crying among depressed participants will be
associated with greater psychosocial impairment. To evaluate
this hypothesis, we examined whether crying within the sample of
depressed individuals was related to severity of depression, length
of depressive episode, and scores on a measure of global functioning. If alterations in the crying response reflected the nature and
extent of emotion dysregulation, crying behavior among depressed
individuals would be related to their level of psychosocial impairment. More specifically, we predicted that among depressed participants, crying during the cry-eliciting film would be positively
associated with depression severity and length and negatively
associated with global functioning.
To test this hypothesis, we conducted t tests comparing crying
and noncrying depressed individuals with respect to their current
self-reported depression severity and global assessment of functioning. Because of the positive skew in length of depressive
episode, we used a chi-square test to examine the association
between episode length and crying. For this analysis, we used a
cutoff of 6 months (the median length for depressive episodes
observed in prior work; Keller et al., 1992) to subdivide the
depressed sample into those with shorter and those with longer
episode duration.
Method
Participants
Seventy-one depressed persons (66% women and 34% men) and 33
nondepressed psychiatrically healthy controls (70% women and 30% men)
participated in the present study. Participants were fluent English speakers
between the ages of 18 and 60 years. Approximately half of the depressed
participants were recruited from two outpatient psychiatry clinics in a
university teaching hospital, and half responded to advertisements posted
in numerous locations in the community (e.g., Internet bulletin boards,
university kiosks, supermarkets). The depressed participants had no reported history of brain injury, no lifetime history of primary psychotic
ideation, no current diagnoses of panic disorder or social phobia, no
F(1, 66) ⫽ 1.21, p ⬎ .05. Our central predictions involved statistical
interactions between cry status and diagnostic status in levels of emotional
reactivity. Crying among the depressed participants was not related to the
number of psychotropic medications used, t(70) ⬍ 1, nor did depressed
psychotropic medication users and nonusers differ in their likelihood of
crying, ␹c2(1, N ⫽ 71) ⫽ 1.59, ⫽ p ⬎ .05. Because medication use was not
related to baseline sympathetic parameters or to primary measures of
emotional reactivity, a statistical interaction of depression status and cry
status would not be confounded by medication use. Nevertheless, because
previous research had found medication effects in depressed samples (e.g.,
Jacobsen, Haukson, & Vestergaard, 1984), we included psychotropic medication status as a covariate in our analyses.
behavioral indications of possible impaired mental status and were not
mentally retarded. Depressed participants were also excluded from the
sample if they were alcohol or substance dependent or if they showed signs
of substance or alcohol abuse within the past 6 months. Thirty-one of the
depressed participants were receiving pharmacotherapy.
The nondepressed nonpsychiatric controls were recruited from the community through advertisements posted in the same locations that were used
to solicit depressed participants. Potential control participants were excluded from the study on the basis of the same general and medical criteria
that were used for the clinical participants. In addition, they were interviewed to exclude those with the presence of lifetime diagnoses of any
Axis I disorder (see below). All participants provided written informed
consent before the experimental session and were paid $25 per hour for
their participation in the study.
Clinical Assessments
All depressed participants met DSM–IV criteria for MDD, as measured
by the Structured Clinical Interview for DSM–IV Axis I Disorders
(SCID–I; First, Gibbons, Spitzer, & Williams, 1995). The SCID–I interviews were conducted by trained advanced graduate students and a postdoctoral fellow. An independent trained rater who was unaware of group
membership evaluated 15 randomly selected audiotapes of SCID–I interviews conducted with depressed and nondepressed participants and with
nonparticipants who met diagnostic criteria for disorders other than depression (e.g., panic disorder). For each interview, the rater evaluated
whether the participant met DSM–IV diagnostic criteria for MDD. In all 15
cases, ratings matched the diagnosis made by the original interviewer
(␬ ⫽ 1.00).
Participants also completed the Beck Depression Inventory (BDI; Beck
et al., 1979) and were rated by the interviewers on the Global Assessment
of Functioning scale (GAF; American Psychiatric Association, 1994).
Finally, for each depressed participant, the SCID–I interviewer determined
the length of the current depressive episode.
Psychophysiology Assessments
Film stimuli. Film stimuli were presented on a 20-inch (50.8-cm)
television monitor at a viewing distance of 1.75 m. Film selection was
based on criteria recommended by Gross and Levenson (1995). All participants viewed two films. A neutral film was used to provide a baseline
against which we could compare the effects of the cry-inducing film. (See
Piferi, Kline, Younger, and Lawler, 2000, for a discussion of the advantages of neutral film baseline procedures.) The neutral film was piloted to
verify that it elicits little emotional behavior or report of emotion. This film
lasted 60 s and depicted coastal landscape scenery. The sad film was
chosen to evoke themes of loss and abandonment that are salient to
depressed individuals (Beck, 1967). This film has been used previously and
is known to elicit strong reports of sadness (e.g., Gross, Sutton, & Ketelaar,
1998, see Gross & Levenson, 1995, for published validation data). This
film was 170 s in length. The film depicted a social scene in which several
individuals grieve in the immediate aftermath of a death. The most poignant figure in the film is a young boy who learns of the death of his father,
reacting first with disbelief, then trying to awaken his dead father and, as
the scene develops, becoming increasingly distraught and inconsolable.
Self-report measures. An emotion inventory taken from Gross et al.
(1998) was used to measure emotion experience at baseline and in response
to the sad film. Each item on the emotion inventory was rated on a 9-point
scale, ranging from 0 (not at all) to 8 (extremely). The inventory included
sadness, which was the emotion targeted by the cry-inducing film. The
inventory also included terms designed to assess the broader dimensions of
positive affect and negative affect (Watson & Tellegen, 1985). A broad
sampling of affective experience was desirable because depression has
been associated with both increased negative affect and decreased positive
CRYING IN DEPRESSION
affect. The Positive Affect scale included six terms: elated, enthusiastic,
amused, happy, lively, and interested (␣ at baseline ⫽ .85). The Negative
Affect scale included six terms: annoyed, anxious, distressed, hostile,
jittery, and nervous (␣ at baseline ⫽ .85).
Behavioral measures. A remote-controlled camera positioned behind
darkened glass unobtrusively recorded the participants’ facial expressive
behaviors. Recording took place in low ambient light. To assess behavioral
reactivity, sadness and crying behaviors were coded using the Emotional
Behavior Coding System (Gross & Levenson, 1993). This system requires
coders to view and rate videotapes in real time. Six undergraduate research
assistants read the training manual for this coding system and participated
in 10 weeks of training before beginning work on this project. All raters
were unaware of the diagnostic status of participants, the nature of the
experimental manipulation, and the study hypotheses. Behavioral coding
training required coders to learn prototypes of emotional behaviors (Ekman
& Friesen, 1975) as well as to recognize specific molecular behaviors
indicative of each prototype. Whereas sadness and crying behaviors are
both manifestations of sadness and would therefore tend to be correlated,
scores for sadness also reflect a variety of molecular behaviors that do not
depend on the presence of crying. Sadness behavior was rated on a 0 – 6
scale, with each value representing an aggregate of intensity, duration, and
frequency of response. For example, a participant would receive a score
of 1 on this scale if she or he exhibited a relatively brief downturning of the
mouth or a slight upturning of the inner eyebrow with closed body posture
and the head moved forward. Participants received a sadness score that
reflected their maximal responding to the neutral film and the sad film. To
ensure that each film score represented comparable time periods, the longer
sad film was divided into two rating periods that were averaged. All
segments of behavior were rated by at least two coders. The average
interrater agreement for observer-rated sadness was acceptable (␣ ⫽ .71).
Crying was coded when the participant displayed visible tears in at least
one eye, not due to yawning. Discrepancies in crying coding were resolved
through consensus discussion after sadness ratings had been completed.
Finally, participants’ latency to cry was measured by recording the time
from the onset of the sad film stimulus to the first appearance of visible
tears.
Physiological measures. An SA Instruments 12-channel bioamplifier
was used to record physiological responses. Data were acquired using a
Pentium PC that used a Data Translation 3001 PCI 12-bit 16-channel
analog-to-digital converter. During the experimental session, physiological
channels were sampled continuously at 400 Hz. Physiological data were
reduced off-line using custom laboratory software written in MATLAB
(Wilhelm, Grossman, & Roth, 1999). Data reduction software interfaced
with recorded binary data files to extract segments of raw data and then
performed waveform transformation, feature detection, and graphic display
for each channel and derived parameter. Artifactual epochs were edited
manually for each channel. A 60-Hz digital notch filter was used to
attenuate ambient electronic noise. Period averages were calculated separately for the 60-s neutral film and the 170-s sad film.
Heart rate, skin conductance level and response, finger temperature,
respiratory rate, and somatic activity were measured. Three criteria guided
selection of these measures: (a) samples from major organ systems known
to be important in emotional responding (cardiac: heart rate; vascular:
finger temperature; electrodermal: skin conductance level and response;
somatic: motor activity; respiratory: respiratory rate); (b) indexes of the
intensity of responding in a sad context (e.g., heart rate acceleration,
Schwartz, Weinberger, & Singer, 1981; electrodermal activation and increases in respiratory rate, Levenson, Ekman, Heider, & Friesen, 1992); (3)
sensitivity to the effects of crying (all measures, Gross et al., 1994; heart
rate acceleration and electrodermal activation, Kraemer & Hastrup, 1988).
The following six measures were obtained:
(a) Heart rate: Beckman miniature electrodes were placed in a bipolar
configuration on opposite sides of the participant’s chest. The interbeat
interval was calculated as the interval (in milliseconds) between successive
305
R waves in the electrocardiogram (ECG) and was converted to instantaneous heart rate after editing of R–R interval outliers due to movement
artifacts or ectopic myocardial activity.
(b) Skin conductance level: A constant-voltage device passed 0.5 V
between Beckman regular electrodes (using an electrolyte of sodium chloride in Unibase) attached to the palmar surface of the proximal phalanges
of the first and second fingers of the nondominant hand. The mean skin
conductance level was analyzed after movement and electrode contact
artifacts appearing as irregular spikes had been edited out using a special
procedure (Wilhelm & Roth, 1996).
(c) Skin conductance response: Skin conductance fluctuations were
detected as changes in skin conductance level from a zero-slope baseline
exceeding 0.2 ␮S.
(d) Finger temperature: A thermistor attached to the palmar surface of
the distal phalange of the fourth finger recorded temperature in degrees
Fahrenheit.
(e) Somatic activity: A piezoelectric sensor attached to the leg of the
participant’s chair provided a sensitive index of overall body movement.
(f) Respiratory rate: Two bands (Respitrace) were placed around the
upper thorax and abdomen and connected to an inductive plethysmography
system (Ambulatory Monitoring, Ardsley, NY).
Procedure
Participants were greeted and were positioned in a comfortable chair
facing a video monitor in a quiet, well-furnished laboratory room. After
procedures for connecting physiological monitoring devices were explained and the monitors were attached, participants performed a number
of experimental tasks that were not relevant to this report. They then were
shown neutral and sad films. Participants’ expressive behavior and physiology were assessed during the neutral and sad films to permit a continuous assessment of changes that accompanied crying. Because pilot data
indicated that assessing self-reported emotional experience between films
disrupted film viewing, participants completed their baseline emotion
questionnaires immediately before they viewed the neutral film. The sad
film was presented directly after the neutral film, and participants’ selfreported responses to the sad film were assessed immediately after they
viewed the sad film. Participants were simply instructed to watch the films
carefully. Finally, to avoid possible carryover effects of the sad film, the
neutral film was always shown before the sad film.
Results
Demographic and Clinical Characteristics
As is evident from the data presented in Table 1, the depressed
and nondepressed participants did not differ with respect to age,
gender composition, and level of education (all ps ⬎ .05). As
expected, the two groups of participants did differ significantly
with respect to severity of depressive symptomatology and level of
global functioning (both ps ⬍ .001).
Manipulation Check
We first sought to determine, for all participants, whether the
sad film affected the report and display of emotions relevant to
crying. As expected on the basis of extensive pretesting, participants reported higher levels of the target emotion of sadness after
viewing the sad film (M ⫽ 4.16, SD ⫽ 2.66) than they reported on
their prefilm emotion inventory (M ⫽ 2.24, SD ⫽ 2.26),
t(103) ⫽ 6.71, p ⬍ .001. The sad film also produced behavior
consistent with a sad emotional state: Participants displayed
greater sadness behavior while viewing the sad film (M ⫽ 1.31,
ROTTENBERG, GROSS, WILHELM, NAJMI, AND GOTLIB
306
Table 1
Sample Characteristics
Age
Education
Group
n
M
SD
Depressed
Nondepressed
71
33
33.4
32.3
10.5
11.7
% Women
66.2%
69.7%
33.8%
30.3%
BDI
GAF
M
SD
M
SD
M
SD
6.5
6.6
1.5
1.4
23.1
2.6
7.4
3.0
87.2
53.7
4.2
8.4
Note. Education ⫽ 1– 8 interval scale, a score of 6.5 was equivalent to part college; BDI ⫽ Beck Depression
Inventory score; GAF ⫽ Global Assessment of Functioning score. BDI and GAF groups differ significantly ( p
⬍.001).
SD ⫽ 1.21) than they did while viewing the neutral film
(M ⫽ 0.19, SD ⫽ 0.61), t(103) ⫽ 9.72, p ⬍ .001. Finally, in
keeping with Gross et al.’s (1994) report of crying frequency, 23
of the 104 participants (23.0%) in this study exhibited observable
crying in response to the sad film.
Hypothesis 1: Likelihood of Crying
Contrary to expectation, depressed and nondepressed participants did not differ significantly from each other in their likelihood
of crying in response to the sad film (23.6% vs. 18.2%, respectively), ␹c2(1, N ⫽ 104) ⬍ 1. Moreover, depressed and nondepressed criers were not distinguished by either their latency of tear
onset (M ⫽ 116.2 s vs. M ⫽ 110.5 s, respectively), t(21) ⬍ 1, or
the length of their crying episodes (M ⫽ 54.7 s vs. M ⫽ 59.5 s,
respectively), t(21) ⬍ 1. The lack of diagnostic group differences
in tearful behavior did not support the hypothesis that major
depression is characterized by a lowered threshold for crying.
As suggested by previous research (see Gross et al., 1994), the
likelihood of crying did differ as a function of gender: Women
were more likely to cry than were men (31% vs. 3%, respectively),
␹c2(1, N ⫽ 104) ⫽ 10.50, p ⬍ .001. Although this pattern was
apparent in both diagnostic groups, our cell sizes did not permit us
to examine whether the magnitude of this gender difference varied
by diagnosis. In the depressed group, 17 of the 48 women cried
(35%) compared with 0 of the 23 men (0%), ␹c2(1, N ⫽
71) ⫽ 11.13, p ⬍ .001; in the nondepressed control group, 5 of
the 24 women cried (21%) compared with 1 of the 9 men (11%),
␹c2(1, N ⫽ 33) ⬍ 1.
Hypothesis 2: Intensity of the Crying Response
Hypothesis 2 predicted that the experiential, behavioral, and
physiological consequences of crying would be enhanced by depression. As we described earlier, in this hypothesis, we predicted
three-way interactions of diagnostic group, cry status, and film
condition for our measures of emotion experience, expressive
behavior, and physiological responding.
Emotion experience. For self-reported sadness, we obtained
the predicted three-way interaction of diagnostic group, cry status,
and film condition, F(7, 97) ⫽ 3.92, p ⫽ .05. Means and standard
deviations of self-reported sadness broken down by diagnostic
group, cry status, and film condition are presented graphically in
Figure 1a. To determine whether depression was associated with
an enhanced crying response, two-way analyses of covariance
(ANCOVAs; Cry Status ⫻ Film Condition, with medication use as
a covariate) were conducted on self-reported sadness separately for
two diagnostic groups. The Cry Status ⫻ Film Condition interaction was not significant for the depressed participants, F(1,
67) ⬍ 1. The fact that the initial levels of sadness reported by the
depressed criers (M ⫽ 4.00, SD ⫽ 2.24) were marginally higher
than those reported by the depressed noncriers (M ⫽ 2.85,
SD ⫽ 2.10), t(69) ⫽ 1.93, p ⬍ .06, may have contributed to the
nonsignificant nature of this interaction. In contrast, the interaction
of cry status and film condition was significant for the nondepressed participants, F(1, 29) ⫽ 8.24, p ⬍ .01. Follow-up analyses
indicated that whereas nondepressed criers and noncriers reported
comparable levels of sadness at baseline, t(31) ⬍ 1, the nondepressed criers reported greater sadness in response to the sad
film than did the nondepressed noncriers, t(31) ⫽ 3.91, p ⬍ .01.
Finally, a change score analysis directly comparing depressed
criers’ and nondepressed criers’ increase in sadness reported to the
sad film yielded comparable results: Depressed criers reported
significantly smaller increases in self-reported sadness to the sad
film (M ⫽ 1.53, SD ⫽ 2.92) than did nondepressed criers
(M ⫽ 5.83, SD ⫽ 0.75), t(21) ⫽ 3.52, p ⬍ .005. These findings for
self-reported sadness, therefore, are inconsistent with the hypothesis that depression enhances the crying response.
The ANCOVA conducted on self-reported levels of negative
affect yielded only a significant main effect for diagnostic group:
Not unexpectedly, depressed individuals (M ⫽ 12.11, SD ⫽ 7.71)
reported more negative affect than did nondepressed individuals
(M ⫽ 6.24, SD ⫽ 4.74) across both cry status and film condition,
F(1, 96) ⫽ 11.33, p ⬍ .001. Finally, the ANCOVA conducted on
levels of positive affect yielded a main effect for film condition,
F(1, 96) ⫽ 64.98, p ⬍ .001, which was subsumed by a significant
interaction of diagnostic group and film condition, F(1,
96) ⫽ 9.86, p ⬍ .005. Follow-up t tests indicated that whereas at
baseline, nondepressed participants reported more positive affect
(M ⫽ 19.61, SD ⫽ 9.88) than did depressed participants
(M ⫽ 12.41, SD ⫽ 6.91), t(100) ⫽ 4.26, p ⬍ .001, the two
diagnostic groups did not differ in their report of positive affect in
response to the sad film (nondepressed: M ⫽ 5.23, SD ⫽ 3.89;
depressed: M ⫽ 6.19, SD ⫽ 5.31), t(100) ⬍ 1. In keeping with this
pattern of findings, the magnitude of the change in positive affect
from baseline to the sad film was greater for the nondepressed than
for the depressed participants, t(100) ⫽ 4.57, p ⬍ .001.
Expressive behavior. As predicted, the ANCOVA conducted
on observer-rated sadness yielded a three-way interaction of cry
status, film condition, and diagnostic group, F(1, 98) ⫽ 4.77, p ⬍
.05. Means and standard deviations of observer-rated sadness,
CRYING IN DEPRESSION
Figure 1. (a) Mean Self-Reported Sadness, (b) mean Observer-Rated Sadness, (c) mean Heart Rate (beats/
min), (d) mean Skin Conductance Level [␮S], (e) mean Skin Conductance Response Rate (response count/
min) ⫻ Cry Status, Film Condition, and Diagnostic Status. Error bars represent standard deviations. In Panel 1a,
baseline refers to responses before the neutral film; in Panels 1b–1e, baseline refers to responses during the
neutral film.
307
308
ROTTENBERG, GROSS, WILHELM, NAJMI, AND GOTLIB
broken down by cry status, film condition, and diagnostic group,
are presented graphically in Figure 1b. To determine whether
depression was associated with an enhanced crying response,
two-way ANCOVAs (Cry Status ⫻ Film Condition, with medication use as a covariate) were conducted on observer-rated sadness
separately for each diagnostic group. The interaction of cry status
and film condition was significant in both diagnostic groups:
nondepressed participants, F(1, 29) ⫽ 47.06, p ⬍ .001; depressed
participants F(1, 68) ⫽ 15.88, p ⬍ .01. Follow-up analyses suggested that these two-way interactions were similar in nature in the
two groups but different in magnitude. Specifically, follow-up t
tests revealed that in the nondepressed group, criers were rated as
sadder than noncriers during the sad film, t(31) ⫽ 8.64, p ⬍ .001,
but not during the neutral film, t(31) ⫽ 1.40, p ⬎ .15. In the
depressed group, criers were rated as sadder than noncriers during
the sad film, t(70) ⫽ 6.15, p ⬍ .001, and marginally so during the
neutral film, t(70) ⫽ 1.70, p ⫽ .08. Contrary to prediction, then,
whether tears were present in response to a sad film had less
impact on the sadness behavior of depressed participants than it
did for nondepressed participants. Finally, a change score analysis
directly comparing depressed criers’ with nondepressed criers’
increases in observable sadness to the sad film yielded conceptually similar, but statistically nonsignificant, results (depressed criers: M ⫽ 2.04, SD ⫽ 1.17; nondepressed criers: M ⫽ 2.92,
SD ⫽ 1.02), t(21) ⫽ 1.62, p ⬎ .05. It is notable that, despite higher
levels of observable sadness among depressed criers during the
neutral film, there is not evidence of a ceiling effect. Indeed, an
inspection of Figure 1b indicates that as expected, the neutral film
elicited low levels of observable sadness across all groups (i.e., all
four cry groups ⬍ 1 on the 0 – 6 scale).
Autonomic physiology. To examine the consequences of crying on autonomic physiology in depressed and nondepressed participants, a repeated measures multivariate analysis of covariance
(MANCOVA) was conducted on the physiological measures. This
analysis also yielded the predicted three-way interaction of cry
status, film condition, and diagnostic group, F(1, 95) ⫽ 8.07, p ⬍
.01. To determine whether depression was associated with an
enhanced crying response, separate two-way MANCOVAs (Cry
Status ⫻ Film Condition, with medication use as a covariate) were
conducted on the six autonomic parameters for the two diagnostic
groups. The interaction of cry status and film condition was not
significant for the depressed participants, F(1, 65) ⬍ 1, but was
significant for the nondepressed participants, F(1, 29) ⫽ 11.30, p
⬍ .005.
To determine which variables contributed to this significant
interaction for the nondepressed participants, we conducted separate two-way ANCOVAs (Cry Status ⫻ Film Condition) on each
of the six physiological measures for this group of participants.
These analyses yielded significant interactions of cry status and
film condition for heart rate, F(1, 29) ⫽ 16.47, p ⬍ .01, skin
conductance level, F(1, 29) ⫽ 5.64, p ⬍ .05, and skin conductance
response rate, F(1, 29) ⫽ 7.01, p ⫽ .01. Means and standard
deviations of these three variables broken down by cry status, film
condition, and diagnostic group are presented graphically in Figures 1c, 1d, and 1e, respectively. Follow-up analyses indicated that
nondepressed participants who cried during the sad film had relatively greater cardiac and electrodermal responses than did nondepressed participants who did not cry. Specifically, while exhib-
iting comparable heart rate and skin conductance response rates
during the neutral film (all ps ⬎ .05), during the sad film nondepressed criers had higher heart rates, t(31) ⫽ 2.87, p ⬍ .01, and a
greater number of skin conductance responses, t(31) ⫽ 2.70, p ⬍
.01, than did nondepressed noncriers. The two groups of nondepressed participants did not differ in skin conductance level for
either the neutral or the sad film, both ts(31) ⬍ .15. Analyses of
change scores conducted to locate the source of the interaction for
skin conductance level indicated that nondepressed noncriers exhibited significant decreases in skin conductance level from the
neutral to the sad film, t(26) ⫽ 2.73, p ⬍ .05, whereas nondepressed criers did not change their skin conductance level between
the two films, t(5) ⫽ 1.43, p ⬎ .05. Thus, whereas nondepressed
criers exhibited greater sympathetic activation during the sad film
than did nondepressed noncriers, among depressed participants,
the presence or absence of observable tears did not have significant
physiological consequences. Inspection of the relevant means in
Table 2 suggests that ceiling or initial level effects on physiological measures assessed during the neutral film are an unlikely
explanation for this pattern of results.
Finally, analyses comparing the change scores of depressed and
nondepressed criers on each of these three autonomic variables
from the neutral to the sad film yielded a convergent pattern of
results. Depressed criers exhibited significantly smaller increases
in skin conductance level (M ⫽ ⫺0.34, SD ⫽ 0.74) and skin
conductance response rate (M ⫽ 1.82, SD ⫽ 2.25) to the sad film
than did nondepressed criers (level: M ⫽ 0.44, SD ⫽ 0.76; rate:
M ⫽ 4.76, SD ⫽ 3.44), t(21) ⫽ 2.29, p ⬍ .05, and t(21) ⫽ 2.37,
p ⬍ .05, respectively. Depressed criers also had nonsignificantly
smaller increases in heart rate to the sad film (M ⫽ 0.20,
SD ⫽ 6.73) than did nondepressed criers (M ⫽ 3.69, SD ⫽ 2.18),
t(20) ⫽ 1.22, p ⬎ .05. These results, therefore, do not support the
hypothesis that the crying response is enhanced by depression.
Hypothesis 3: Clinical Significance of Crying
Before assessing the clinical significance of crying in depression, we examined the associations between crying and demographic variables. Although depressed criers did not differ significantly from depressed noncriers in age or level of education, both
ts(70) ⬍ 1, there was a strong gender difference in crying (as noted
previously, depressed men did not cry). Because of this gender
difference, we restricted subsequent analyses examining the clinical significance of crying in depression to female depressed
participants.
We predicted that the presence of crying behavior would be
related to more impaired psychosocial functioning among depressed individuals, as measured by self-reported severity of depression, length of the current depressive episode, and global
assessment of functioning scores. The results of the analyses
indicated that depressed participants who cried did not differ from
depressed participants who did not cry, with respect to either
self-reported depression severity, as measured by the BDI, or
global functioning, as measured by the GAF, both ts(46) ⬍ 1, p ⬎
.05. In terms of episode length, a significantly smaller proportion
of depressed women with longer depressive episodes cried (7 of
28; 25.0%) than was the case among depressed women with
CRYING IN DEPRESSION
309
Table 2
Means (and Standard Deviations) of Autonomic Nervous System Measures by Cry Status, Film Condition, and Diagnostic Status
HR
Group
Depressed criers
Neutral film
Sad film
Depressed noncriers
Neutral film
Sad film
Nondepressed criers
Neutral film
Sad film
Nondepressed noncriers
Neutral film
Sad film
SCL
SCR
ACT
TMP
RR
M
SD
M
SD
M
SD
M
SD
M
SD
M
SD
81.81
82.01
9.34
12.82
7.68
7.35
4.71
4.70
2.39
4.21
2.70
2.96
0.89
1.28
0.88
2.88
89.59
88.90
6.12
6.69
13.86
15.05
5.36
2.66
76.31
74.90
12.71
12.74
6.76
6.53
2.86
2.74
2.58
4.14
2.90
2.32
1.30
0.91
2.73
1.02
89.22
88.75
5.97
5.97
14.89
17.46
3.87
2.58
78.09
81.78
10.84
10.54
8.25
8.70
6.21
6.76
4.24
9.00
3.29
5.88
0.86
0.99
0.48
0.45
94.12
94.02
1.77
1.92
13.75
15.32
2.58
2.66
74.11
71.82
7.76
7.03
7.45
7.03
2.96
2.98
2.96
4.36
2.00
3.27
1.20
0.80
1.70
1.12
90.28
89.66
5.34
5.57
13.57
16.80
2.77
3.31
Note. HR ⫽ heart rate (beats/min); SCL ⫽ skin conductance level (␮S); SCR ⫽ skin conductance response count (count/min); ACT ⫽ somatic activity
(A/D units); TMP ⫽ finger temperature (°F); RR ⫽ respiratory rate (breaths/min). There was a significant three-way interaction between HR, SCL, and
SCR ( p ⬍ .05).
shorter episodes (10 of 19; 52.6%), ␹2(1, N ⫽ 47) ⫽ 3.74, p ⬍ .05,
a pattern opposite to that predicted by Hypothesis 3.
Discussion
In this study, we evaluated how clinical depression affects the
crying response. On the basis of clinical reports, we expected that
depressed participants would exhibit tearful behavior more readily
than would nondepressed persons and that depression would enhance the crying response. We also predicted that the presence of
tearful crying among the depressed participants would be associated with longer and more severe depressive episodes and with
greater psychosocial impairment. None of these predictions were
supported. Depressed participants were no more likely to cry than
were their nondepressed counterparts. Moreover, among depressed
individuals, the presence or absence of crying had relatively little
impact on reports or displays of sadness, or on levels of sympathetic arousal. Finally, crying among depressed participants was
not related to severity of depression and was in fact associated with
shorter, not longer, depressive episodes.
Why Didn’t Depression Influence Crying Threshold?
Our finding that there was no difference in crying threshold
between depressed and nondepressed participants was particularly
surprising. We expected that the depressed participants would
begin the experiment in a sad, distressed state and would therefore
be predisposed to cry relatively easily. In fact, our finding of no
difference in crying threshold between the depressed and nondepressed participants was all the more surprising given that the film
we used to elicit sadness depicted themes of relationship loss,
death, and abandonment—themes that are strongly salient to currently depressed persons (Beck, 1983). Despite all of these considerations, no differences in crying likelihood emerged. How
might our findings be reconciled with the observations in clinical
contexts of notable crying among depressed individuals?
One possible reconciliation of this discrepancy is that excessive
crying in depression is observable primarily in social situations. It
is well established, for example, that depressed persons are prone
to disclose and dramatize their negative experiences to others (e.g.,
Coyne, 1976). Indeed, out of this concern, we chose to examine the
threshold for crying behavior outside of the influence of an interpersonal context. Perhaps we would have observed increased tearfulness in depressed persons had we used an interactional paradigm rather than testing each participant alone. The results of the
present study raise the possibility that variations in social context
may be an important determinant of crying in depression, and
future research might benefit from a systematic examination of this
issue.
A second possibility is that there are, in fact, changes in the
threshold for crying in depression but that these changes occur
only in a subgroup of depressed individuals. Indeed, although no
evidence of a lowered crying threshold was obtained in group level
analyses, internal analyses within the sample of depressed persons
suggested that high levels of sadness may nevertheless facilitate
crying in depression. At a trend level, depressed persons who cried
in response to a sad film reported more sadness and looked sadder
before viewing the sad film than did their noncrying counterparts.
Thus, although generally elevated sadness does not appear to
facilitate crying in depression, extremely high levels of sadness
may do so (Wallbott & Scherer, 1988).
Gender Differences in Crying
We also found that women were more likely than men to cry
among both depressed and nondepressed participants. These strong
gender differences in observable tearfulness were found in the
absence of other gender differences in emotional responding to the
sad film. Similar results were obtained by Delp and Sackheim
(1987), who measured lacrimal flow directly while participants
engaged in sad imagery. These investigators found that although
women reported levels of dysphoric affect comparable to those of
men, they demonstrated greater increases in lacrimal flow. Although there may be a biological basis for gender differences in
psychogenic crying, note that developmental research has gener-
310
ROTTENBERG, GROSS, WILHELM, NAJMI, AND GOTLIB
ally found either no sex differences or more frequent and spontaneous crying behavior in boys relative to girls (e.g., Maccoby &
Feldman, 1972). A more comprehensive examination of the developmental origins of gender differences in crying would help to
clarify this issue.
It is important to underscore here that strong gender differences
in crying behavior were observed in both diagnostic groups. Although it would clearly be desirable in future work to ensure that
crying is elicited in both males and females to examine the consequences of crying in both genders, it is noteworthy that gender
differences in crying were preserved in the depressed state. This
pattern is consistent with previous research indicating that compared with depressed females, depressed males are characterized
by attenuated outward expression of sad emotion (e.g., Hammen &
Padesky, 1977; Vredenburg et al., 1986). In the current study,
however, gender differences were observed only with respect to
tear presence and not on a measure of overall sadness behavior.
The present findings clearly highlight the need for future research
to examine more explicitly which aspects of sadness behavior
distinguish depressed males from depressed females and to elucidate more systematically the nature of gender differences in the
clinical presentation of depression.
Depression, Crying Intensity, and Psychosocial
Functioning
As expected on the basis of previous research (e.g., Gross et al.,
1994; Kraemer & Hastrup, 1988), normal (i.e., nondepressed)
participants who cried in response to a sad film exhibited increases
in the report and expression of sadness and showed greater physiological arousal than did nondepressed participants who did not
cry. Contrary to our expectations, however, these differences in
emotional responding between criers and noncriers were less pronounced among diagnosably depressed individuals than they were
among the nondepressed participants. Indeed, the presence or
absence of a tearful response to a sad film had relatively little
impact on depressed participants’ report of sadness, displays of
sadness, or their levels of sympathetic arousal. Data concerning the
impact of the duration of the current depressive episode on observable crying behavior also suggested that being depressed may
serve to blunt the crying response: Those individuals who had been
depressed the longest were the least likely to cry.
How might we account for depression-related attenuation of the
crying response? One possibility is that these findings concerning
the crying response reflect statistical or methodological artifacts
that resulted from differences between our participant groups in
their initial levels of sadness (e.g., Berntson, Uchino, & Cacioppo,
1994). More specifically, it is possible that we obtained a significant three-way interaction in our analysis of self-reported sadness
because depressed criers may have been near ceiling before the sad
film, thereby truncating the possible range of change on selfreported sadness as a function of crying. It is worth noting,
however, that the other two domains of emotional responding
yielded convergent results while being less vulnerable to initiallevels explanations. In fact, group differences in initial levels are
an unlikely explanation for the three-way interactions obtained
with both the behavioral and the autonomic data. There was no
evidence that depressed criers were near ceiling on any of these
measures during the neutral film, nor was there evidence of an
association between initial levels and subsequent change in either
diagnostic group. (See Myrtek and Foerster, 1986, for a critique of
the concept of initial level dependency.) Finally, these statistical
considerations cannot explain why longer episodes of depression
were associated with a reduced likelihood of crying, because this
analysis was conducted entirely within the group of depressed
participants.
A more substantive explanation for why the crying response
might become less likely and less powerful over time among
depressed individuals (rather than more likely and more powerful,
as we had hypothesized) involves the prolonged self-regulatory
difficulties that typically occur in depression. Depressed individuals, virtually by definition, have a persistent inability to regulate
their negative affect (Gross & Muñoz, 1995). Indeed, not only does
depression often run a chronic course (e.g., Boland & Keller, in
press; Mueller et al., 1996) but there is also evidence to indicate
that most depressed individuals experience unrelieved negative
affect before the onset of diagnosable episodes of the disorder
(Watson & Walker, 1996). Equally important, depression has a
corrosive effect on socially supportive relationships (Joiner, in
press). Thus, these features of depression suggest that depressed
individuals may be discouraged from mounting vigorous crying
episodes because of a reduced probability that crying episodes will
secure effective help from others or will lead to affect remediation.
Clearly, documenting these implications of the present data will be
a high priority for future research.
The attenuated crying response in depression also may reflect a
broader deficit with respect to emotional functioning in depression.
Typically, emotional responses show some degree of coordination,
enabling an organism to rapidly mobilize an appropriate response
to meaningful stimuli in the environment (Keltner & Gross, 1999;
Plutchik, 1980). Indeed, this rapid mobilization was apparent in the
well-orchestrated changes in experience, behavior, and physiology
that accompanied crying in the nondepressed participants. In contrast, tearfulness was not associated with dramatic changes across
systems of emotional responding among depressed persons. In this
respect, tearful behavior among depressed persons appeared to be
largely disconnected from its usual sequelae. Indeed, the present
results are consistent with previous findings indicating low coherence or agreement across domains of emotional responding in
depressed persons (Allen, Trinder, & Brennen, 1999; Brown,
Schwartz, & Sweeney, 1978). In extending this work and moving
toward a broader account of emotional dysregulation in MDD, it
will be critical that other emotional responses receive the scrutiny
we applied to crying. The findings of the present investigation
should serve notice that the integration of basic affective science
with research on psychopathology may bear important, and surprising, insights regarding the ways in which emotion-related
processes are affected by various forms of disorder.
References
Ainsworth, M. D. S. (1969). Object relations, dependency and attachment:
A theoretical view of the mother–infant relationship. Child Development, 40, 969 –1025.
Allen, N. B., Trinder, J., & Brennen, C. (1999). Affective startle modulation in clinical depression: Preliminary findings. Biological Psychiatry, 46, 542–550.
American Psychiatric Association. (1994). Diagnostic and statistical man-
CRYING IN DEPRESSION
ual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association.
Averill, J. R. (1968). Grief: Its nature and significance. Psychological
Bulletin, 70, 721–748.
Barlow, D. H. (1988). Disorders of emotion. Psychological Inquiry, 2,
58 –71.
Beck, A. T. (1967). Depression: Clinical, experimental, and theoretical
aspects. New York: Harper & Row.
Beck, A. T. (1983). Cognitive therapy of depression: New perspectives. In
P. J. Clayton & J. E. Barrett (Eds.), Treatment of depression: Old
controversies and new approaches (pp. 265–290). New York: Raven
Press.
Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive
therapy of depression. New York: Guilford Press.
Berntson, G. G., Uchino, B. N., & Cacioppo, J. T. (1994). Origins of
baseline variance and the law of initial values. Psychophysiology, 31,
204 –210.
Boland, R. J., & Keller, M. B. (in press). Course and outcome of depression. In I. H. Gotlib & C. L. Hammen (Eds.), Handbook of depression.
New York: Guilford Press.
Bowlby, J. (1969). Attachment and loss: Vol. 1. Attachment. New York:
Basic Books.
Brown, S., Schwartz, G. E., & Sweeney, D. R. (1978). Dissociation of
self-reported and observed pleasure in depression. Psychosomatic Medicine, 40, 536 –548.
Carter, J. D., Joyce, P. R., Mulder, R. T., Luty, S. E., & McKenzie, J.
(2000). Gender differences in the presentation of depressed outpatients:
A comparison of descriptive variables. Journal of Affective Disorders, 61, 59 – 67.
Choti, S. E., Marston, A. R., Holston, S. G., & Hart, J. L. (1987). Gender
and personality variables in film-induced sadness and crying. Journal of
Social and Clinical Psychology, 5, 435–544.
Cornelius, R. R. (1984). A rule model of adult emotional expression. In
C. Z. Malatesta & C. E. Izard (Eds.), Emotion in adult development (pp.
213–233). Beverly Hills, CA: Sage.
Cornelius, R. R. (1988, April). Toward an ecological theory of crying and
weeping. Paper presented at the meeting of the Eastern Psychological
Association, Buffalo, NY.
Coyne, J. C. (1976). Toward an interactional description of depression.
Psychiatry, 39, 28 – 40.
Darwin, C. (1872). The expression of the emotions in animals and man.
New York: Appleton.
Davis, D., Lambert, J., & Ajans, Z. A. (1969). Crying in depression. British
Journal of Psychiatry, 115, 597–598.
Delp, M. J., & Sackheim, H. A. (1987). Effects of mood on lacrimal flow:
Sex differences and asymmetry. Psychophysiology, 24, 550 –556.
Efran, J. S., & Spangler, T. J. (1979). Why grown-ups cry: A two factor
theory and evidence from The Miracle Worker. Motivation and Emotion, 3, 63–72.
Ekman, P., & Friesen, W. V. (1975). Unmasking the face. Englewood
Cliffs, NJ: Prentice Hall.
First, M. B., Gibbon, M., Spitzer, R. L., & Williams, J. B. W. (1995).
User’s guide for the Structured Clinical Interview for DSM–IV Axis I
disorders (SCID-I, Version 2.0). Washington, DC: American Psychiatric
Press.
Frey, W. H., II, DeSota-Johnson, D., Hoffman, C., & McCall, J. T. (1981).
Effect of stimulus on the chemical composition of human tears. American Journal of Ophthalmology, 92, 559 –567.
Gotlib, I. H., & Neubauer, D. L. (2000). Information-processing approaches to the study of cognitive biases in depression. In S. L. Johnson,
A. M. Hayes, T. M. Field, N. Schneiderman, & P. M. McCabe (Eds.),
Stress, coping, and depression (pp. 117–143). Mahwah, NJ: Erlbaum.
Gross, J. J., Fredrickson, B. L., & Levenson, R. W. (1994). The psychophysiology of crying. Psychophysiology, 31, 460 – 468.
311
Gross, J. J., & Levenson, R. W. (1993). Emotional suppression: Physiology, self-report, and expressive behavior. Journal of Personality and
Social Psychology, 64, 970 –986.
Gross, J. J., & Levenson, R. W. (1995). Emotion elicitation using films.
Cognition and Emotion, 9, 81–108.
Gross, J. J., & Muñoz, R. F. (1995). Emotion regulation and mental health.
Clinical Psychology: Science and Practice, 2, 151–164.
Gross, J. J., Sutton, S. K., & Ketelaar, T. (1998). Relations between affect
and personality: Support for the affect-level and affective reactivity
views. Personality and Social Psychology Bulletin, 24, 279 –288.
Hamilton, M. (1960). A rating scale for depression. Journal of Neurology,
Neurosurgery and Psychiatry, 12, 56 – 62.
Hammen, C. L., & Padesky, C. A. (1977). Sex differences in the expression
of depressive responses on the Beck Depression Inventory. Journal of
Abnormal Psychology, 86, 609 – 614.
Hunziker, U., & Barr, B. (1986). Increased carrying reduces infant crying.
Pediatrics, 77, 641– 648.
Jacobsen, J., Haukson, P., & Vestergaard, P. (1984). Heart rate variation in
patients treated with antidepressants. An index of anticholinergic effects? Psychopharmacology, 84, 544 –548.
Joiner, T. E., Jr. (in press). Depression in its interpersonal context. In I. H.
Gotlib & C. L. Hammen (Eds.), Handbook of depression. New York:
Guilford Press.
Keller, M. B., Lavori, P. W., Mueller, T. I., Endicott, J., Coryell, W.,
Hirschfeld, R. M. A., & Shea, T. (1992). Time to recovery, chronicity,
and levels of psychopathology in major depression: A 5-year prospective
follow-up of 431 subjects. Archives of General Psychiatry, 49, 809 –
816.
Keltner, D., & Gross, J. J. (1999). Functional accounts of emotions.
Cognition and Emotion, 13, 467– 480.
Kraemer, D. L., & Hastrup, J. L. (1986). Crying in natural settings: Global
estimates, self-monitored frequencies, depression and sex differences in
an undergraduate population. Behavior Research and Therapy, 24, 371–
373.
Kraemer, D. L., & Hastrup, J. L. (1988). Crying in adults: Self-control and
autonomic correlates. Journal of Social and Clinical Psychology, 6,
53–58.
Kring, A. M., & Bachorowski, J. (1999). Emotions and psychopathology.
Cognition and Emotion, 13, 575–599.
Labott, S. M., & Martin, R. B. (1987). The stress-moderating effects of
weeping and humor. Journal of Human Stress, 13, 159 –164.
Levenson, R. W., Ekman, P., Heider, K., & Friesen, W. V. (1992). Emotion
and autonomic nervous system activity in the Minangkabau of West
Sumatra. Journal of Personality and Social Psychology, 62, 972–988.
Maccoby, E. E., & Feldman, S. S. (1972). Mother-attachment and strangerreactions in the third year of life. Monographs of the Society for
Research in Child Development, 37, 86.
Maclean, P. (1985). Brain evolution relating to family, play, and the
separation call. Archives of General Psychiatry, 442, 405– 422.
Mueller, T. I., Keller, M. B., Leon, A. C., Solomon, D. A., Shea, M. T.,
Coryell, W., & Endicott, J. (1996). Recovery after 5 years of unremitting
major depressive disorder. Archives of General Psychiatry, 53, 794 –
799.
Myrtek, M., & Foerster, F. (1986). The law of initial value: A rare
exception. Biological Psychology, 22, 227–237.
Newman, J. (1985). The infant cry of primates: An evolutionary perspective. In B. Lester & C. Boukydis (Eds.), Infant crying (pp. 307–323).
New York: Plenum.
Piferi, R. L., Kline, K. A., Younger, J., & Lawler, K. A. (2000). An
alternative approach for achieving cardiovascular baseline: Viewing an
aquatic video. International Journal of Psychophysiology, 37, 207–217.
Plutchik, R. (1980). Emotion, a psychoevolutionary synthesis. New York:
Harper & Row.
Post, R. M., & Weiss, S. R. B. (1998). Sensitization and kindling phenom-
312
ROTTENBERG, GROSS, WILHELM, NAJMI, AND GOTLIB
ena in mood, anxiety, and obsessive– compulsive disorders: The role of
serotonergic mechanisms in illness progression. Biological Psychiatry, 44, 193–206.
Schwartz, G. E., Weinberger, D. A., & Singer, J. A. (1981). Cardiovascular
differentiation of happiness, sadness, anger, and fear following imagery
and exercise. Psychosomatic Medicine, 43, 343–364.
Tompkins, S. S. (1963). Affect, imagery, consciousness (Vol. 2). New
York: Springer Press.
Vredenburg, K., Krames, L., & Flett, G. (1986). Sex differences in the
clinical expression of depression. Sex Roles, 14, 37– 49.
Wallbott, H. G., & Scherer, K. R. (1988). How universal and specific is
emotional experience? Evidence from 27 countries on five continents. In
K. R. Scherer (Ed.), Facets of emotion: Recent research (pp. 31–56).
Hillsdale, NJ: Erlbaum.
Watson, D., Clark, L. A., & Carey, G. (1988). Positive and negative
affectivity and their relation to anxiety and depressive disorders. Journal
of Abnormal Psychology, 97, 346 –353.
Watson, D., & Tellegen, A. (1985). Towards a consensual structure of
mood. Psychological Bulletin, 98, 219 –235.
Watson, D., & Walker, L. M. (1996). The long-term stability and predictive
validity of trait measures of affect. Journal of Personality and Social
Psychology, 70, 567–577.
Wilhelm, F. H., Grossman, P., & Roth, W. T. (1999). Analysis of cardiovascular regulation. Biomedical Science Instrumentation, 35, 135–140.
Wilhelm, F. H., & Roth, W. T. (1996). Ambulatory assessment of clinical
anxiety. In J. Fahrenberg (Ed.), Ambulatory assessment: Computerassisted psychological and psychophysiological methods in monitoring
and field studies (pp. 317–345). Göttingen, Germany: Hogrefe & Huber.
Williams, D. G. (1982). Weeping by adults: Personality correlates and sex
differences. Journal of Psychology, 110, 217–226.
Zetin, M., Sklansky, G. J., & Cramer, M. (1984). Sex differences in
inpatients with major depression. American Journal of Psychiatry, 45,
257–259.
Received May 21, 2001
Revision received September 20, 2001
Accepted September 25, 2001 䡲